Intel Iris Pro Graphics 6200 vs NVIDIA GeForce GTX 670M Comparison

The NVIDIA GeForce GTX 670M and Intel Iris Pro Graphics 6200 represent two fundamentally different approaches to mobile graphics, separated by two years and a massive shift in process technology. The benchmark data shows a clear single winner in compute workloads, but the full picture requires examining architecture, power, and feature support. The GTX 670M leads in the only shared benchmark, yet the Iris Pro offers a higher peak compute figure and modern API coverage, making the choice dependent on the user's priorities.

Head-to-Head Benchmarks

The sole head-to-head comparison available is the Geekbench OpenCL test, which measures general-purpose compute performance. In this test, the NVIDIA GeForce GTX 670M scores 6513 points, while the Intel Iris Pro Graphics 6200 scores 4556 points. This gives the GTX 670M a decisive 43% advantage over the Intel part. That is a substantial margin, indicating that for compute-heavy tasks such as OpenCL-accelerated video encoding, physics simulations, or data processing, the NVIDIA solution delivers significantly more throughput.

Contextualizing these scores against their respective rivals reinforces this gap. The GTX 670M sits at the 38th percentile of all GPUs, with an average benchmark score of 6513. Its nearest rival, the NVIDIA GeForce GT 555M, scores 6493, putting the GTX 670M just 0.3% ahead. The AMD Radeon Vega 10 Mobile is close behind at 6476 (0.6% slower), and the Intel UHD Graphics P750 scores 6554, which is 0.6% faster than the GTX 670M. This means the GTX 670M is competitive with modern integrated graphics, despite being older hardware.

The Intel Iris Pro Graphics 6200, by contrast, sits at the 35th percentile with an average score of 6117. Its OpenCL score of 4556 is notably lower than its average, dragging down its overall standing. Its nearest rival, the AMD Radeon HD 8690M, scores 6137, which is 0.3% faster. The NVIDIA RTX A400 scores 6078 (0.6% slower), the GeForce MX230 scores 6077 (0.7% slower), and the Quadro P2000 scores 6049 (1.1% slower). Interestingly, the Iris Pro's average score is higher than its OpenCL score because it also runs Metal and Vulkan benchmarks, where it performs much better — 7764 in Metal and 6032 in Vulkan. These figures show that the Iris Pro is a more versatile part in terms of API support, but in raw OpenCL, it lags significantly behind the GTX 670M.

The delta of 43% in OpenCL is the defining performance metric between these two. For a user who primarily runs OpenCL workloads, the GTX 670M is the clear winner. However, the Iris Pro's Metal score of 7764 is higher than the GTX 670M's OpenCL score, suggesting that on platforms and applications that use Metal, the Intel part could outperform the older NVIDIA GPU. This highlights the importance of API selection in benchmarking.

Architecture Differences

The architectural chasm between these two GPUs is vast. The NVIDIA GeForce GTX 670M uses the GF114 chip, based on the Fermi 2.0 architecture, fabricated on a 40 nm process at TSMC. This is a mature design from the GeForce 600M generation, released in March 2012. The chip contains 1,950 million transistors on a 332 mm² die, yielding a transistor density of 5.9 million transistors per mm². This is a large, power-hungry discrete GPU designed for gaming laptops.

The Intel Iris Pro Graphics 6200, on the other hand, uses the Broadwell GT3e chip, based on Intel's Generation 8.0 architecture, fabricated on a 14 nm process at Intel. This is a much newer and denser process, though Intel does not list transistor count or die size for this part. It was released in September 2014, two and a half years after the GTX 670M. The Iris Pro is an integrated graphics processor (IGP) that shares system memory, whereas the GTX 670M has its own dedicated GDDR5 memory.

Memory configurations could not be more different. The GTX 670M has 1536 MB of GDDR5 memory on a 192-bit bus, delivering 72.00 GB/s of bandwidth. The memory clock is 750 MHz, or 3 Gbps effective. The Iris Pro uses System Shared memory, with a System Shared bus width and System Dependent bandwidth. This means the Intel part's memory performance is entirely dependent on the system's RAM, which is typically slower than dedicated GDDR5 but can be larger.

Core counts also differ. The GTX 670M has 336 shading units, 56 texture mapping units (TMUs), and 24 raster output units (ROPs). The Iris Pro has more shading units at 384, but fewer TMUs at 48 and far fewer ROPs at just 6. This explains the pixel rate difference: the GTX 670M achieves 8.372 GPixel/s, while the Iris Pro manages only 6.600 GPixel/s. However, the Iris Pro has a higher texture rate at 52.80 GTexel/s, compared to 33.49 GTexel/s for the GTX 670M. This is because the Intel part has a higher boost clock of 1100 MHz (base 300 MHz), while the NVIDIA part's base and boost clocks are not listed.

In terms of raw floating-point performance, the Iris Pro edges out the GTX 670M. The Intel part delivers 844.8 GFLOPS of FP32 compute, while the NVIDIA part delivers 803.7 GFLOPS. This is a 5% advantage for Intel, which is interesting given the GTX 670M's huge win in OpenCL. This suggests that the NVIDIA architecture is more efficient at extracting real-world performance from its theoretical peak, or that the OpenCL driver and implementation for the GTX 670M is more mature.

Power consumption is another major differentiator. The GTX 670M has a TDP of 75 W, which is typical for a discrete mobile GPU of its era. The Iris Pro has a TDP of just 15 W, making it far more power-efficient. This is a 60 W difference, which has massive implications for battery life and thermal management in laptops. The GTX 670M uses an MXM Module slot width with no power connectors, while the Iris Pro is an IGP on a Ring Bus interface.

The Verdict

The data paints a clear picture: the NVIDIA GeForce GTX 670M is the superior choice for compute-intensive OpenCL workloads, beating the Intel Iris Pro Graphics 6200 by 43% in the head-to-head benchmark. If your primary use case involves applications that leverage OpenCL, the GTX 670M is the only rational pick. Its 75 W TDP and dedicated GDDR5 memory provide the bandwidth and throughput needed for demanding tasks.

However, the Intel Iris Pro Graphics 6200 is not without merit. It has a higher FP32 peak (844.8 GFLOPS vs 803.7 GFLOPS) and a higher texture rate (52.80 GTexel/s vs 33.49 GTexel/s). More importantly, it supports a wider range of modern APIs, including Vulkan 1.0 and DirectX 12 (11_1), whereas the GTX 670M only supports DirectX 12 (11_0) and has no Vulkan support listed. The Iris Pro also scores much higher in Metal (7764) and Vulkan (6032) benchmarks, indicating better performance in those specific environments.

For a user who values power efficiency and modern API support over raw OpenCL compute, the Iris Pro is the better choice. Its 15 W TDP is a fraction of the GTX 670M's 75 W, making it ideal for thin-and-light laptops where battery life is paramount. The GTX 670M, being an MXM module, requires a larger chassis and more cooling, which is acceptable for a gaming laptop but not for an ultraportable.

FAQ

Q: Which GPU is faster in OpenCL?

A: The NVIDIA GeForce GTX 670M is 43% faster than the Intel Iris Pro Graphics 6200 in Geekbench OpenCL, scoring 6513 versus 4556.

Q: Does the Intel Iris Pro Graphics 6200 support Vulkan?

A: Yes, the Intel Iris Pro Graphics 6200 supports Vulkan 1.0, while the NVIDIA GeForce GTX 670M has no Vulkan support listed.

Q: What is the TDP difference between the two?

A: The NVIDIA GeForce GTX 670M has a TDP of 75 W, while the Intel Iris Pro Graphics 6200 has a TDP of 15 W, a difference of 60 W.

Q: Which GPU has more shading units?

A: The Intel Iris Pro Graphics 6200 has 384 shading units, compared to 336 on the NVIDIA GeForce GTX 670M.

Q: What is the memory bandwidth of the GTX 670M?

A: The NVIDIA GeForce GTX 670M has 72.00 GB/s of memory bandwidth, using 1536 MB of GDDR5 on a 192-bit bus.

Q: How do the average benchmark scores compare?

A: The NVIDIA GeForce GTX 670M has an average benchmark score of 6513, which is higher than the Intel Iris Pro's average of 6117. The GTX 670M sits at the 38th percentile, while the Iris Pro sits at the 35th percentile.

Where Each One Wins

The NVIDIA GeForce GTX 670M wins in OpenCL compute performance. Its 6513 score versus 4556 for the Intel part is the single largest benchmark delta in this comparison. It also wins in pixel rate (8.372 GPixel/s vs 6.600 GPixel/s) and has dedicated memory with a fixed 72.00 GB/s bandwidth, which is more predictable than system-shared memory. For tasks like video transcoding, scientific computing, or any OpenCL-accelerated workload, the GTX 670M is the clear winner.

The Intel Iris Pro Graphics 6200 wins in FP32 peak performance (844.8 GFLOPS vs 803.7 GFLOPS) and texture rate (52.80 GTexel/s vs 33.49 GTexel/s). It also wins in API support, offering Vulkan 1.0 and DirectX 12 (11_1) versus the GTX 670M's DirectX 12 (11_0) only. In Metal and Vulkan benchmarks, the Iris Pro scores 7764 and 6032 respectively, which are higher than the GTX 670M's OpenCL score of 6513. For users on macOS or Linux systems that leverage Vulkan, the Iris Pro is the better performer.

Specification Differences

The key specification differences are stark. The process node differs dramatically: the GTX 670M uses 40 nm TSMC, while the Iris Pro uses 14 nm Intel. Transistor count and die size are only listed for the GTX 670M (1,950 million and 332 mm² respectively), with no data for the Intel part. The GTX 670M has a transistor density of 5.9M / mm², which is not listed for the Iris Pro.

Clock speeds differ: the Iris Pro has a base clock of 300 MHz and a boost clock of 1100 MHz, while the GTX 670M has no base or boost clock listed, only a memory clock of 750 MHz (3 Gbps effective). Memory configuration is completely different: the GTX 670M has 1536 MB GDDR5 on a 192-bit bus with 72.00 GB/s bandwidth, while the Iris Pro uses System Shared memory with System Dependent bandwidth.

Core counts differ: the GTX 670M has 336 shading units, 56 TMUs, and 24 ROPs, while the Iris Pro has 384 shading units, 48 TMUs, and 6 ROPs. Pixel rate is higher on the GTX 670M (8.372 GPixel/s vs 6.600 GPixel/s), but texture rate is higher on the Iris Pro (52.80 GTexel/s vs 33.49 GTexel/s). FP32 compute is slightly higher on the Iris Pro (844.8 GFLOPS vs 803.7 GFLOPS).

Power and interface differ: the GTX 670M has a TDP of 75 W and uses an MXM Module slot with MXM-B (3.0) bus interface, while the Iris Pro has a TDP of 15 W and is an IGP on a Ring Bus. Display outputs are Portable Device Dependent for the GTX 670M and Motherboard Dependent for the Iris Pro. API support differs in Vulkan (none for GTX 670M, 1.0 for Iris Pro) and DirectX (12 (11_0) vs 12 (11_1)), while OpenGL is higher on the GTX 670M (4.6 vs 4.4).